The problem of Kalifat and Kumkale

Two small villages, Kalifat and Kumkale appear in the plain of Troy on old maps. 

There should be no villages in a flood plain, because they will be wiped out by flooding.  

Yet these two villages lasted in the plain at least a hundred years, from the 1840s to the 1950s, after which they were moved onto the prominences overlooking their old locations.  

I suspect that those two villages could endure in the plain only because they were on high ground.  

This is clearly the case with old Kumkale.  It sat on the northern end of what Walter Leaf thought was the Throsmos, or swelling of the plain, mentioned by Homer.  The location is marked on Leaf's map below, just northwest of Troia.

Below is a map from Heinrich Schliemann.  It places Kum-Kioi where Leaf puts the throsmos.  Nearby is the Burial mound of Illus.

Below is a photo looking northwest over the top of Troia, toward the mound Leaf had in mind.  The mound is visibly rising at the top of the frame, where a canal goes around it.  Nearer to Troy a road climbs the mound from right to left.  

Below is a more recent view to the northwest from Troia.  Again, the rise Leaf was talking about is visible where the canal goes around it on the right and the road climbs the mound in the center. Out beyond the canal is the new village of Kumkale. 

Below is another view to the northwest.  The rising land below the "corbis" watermark in this photo is pretty dramatic.  At the top of the photo you can see the canal going around the end of the mound. 

Below is yet another pic looking northwest across the excavated area at Hissarlik. The new village of Kumkale is clearly visible on the prominence beyond the canal which is curving around the obvious mound on which it formerly sat. A road climbs the mound from right to left.  

Below is another map from Heinrich Schliemann.  It places Kum-keui near two other place names, Ilus and Polion.  It's as if there were three distinct villages there.  

There can be little doubt that the villages in that area were on top of a mound.  The layout and curvature of the roads on Schliemann's map make the huge mound in front of Hisarlik very distinct. 

The original location of Kalafat is harder to discern.  It sat near the pointed, southern end of the great mound in front of Hissarlik. As long as it was on that mound, it would have had some height above the floor of the valley.  Schliemann's map above places it at the meeting place of two roads. The map below places "Kalifatlee" in a similar location.  

The map below seems to agree on the location of old Kalafat. 

The map from Thomas Spratt below also puts Kalifatli on the edge of the winter channel of the Scamander. 

Finally, on the map below we have Eski (old) Kalafat Koy (village) Yeri (location) marked right next to the Scamander.  This map places the original location of the village in what Spratt called the winter channel in which the Mendere/Scamander runs today.  

I think it is easy to say why Kumkale survived in the plain (it was on a mound) but I cannot say why Kalafat survived in the plain, because it does not appear to have been on the mound.  It appears to have been either in the winter channel or right next to it.  

If Kalafat was not on a mound, then how did it survive floods? 

Below is a modern satellite photo with the approximate locations of the old villages marked.

The two villages used to sit at opposite ends of the great mound in front of Hissarlik.  Whether Kalafat was actually on the mound is questionable.  Perhaps part of it was.  Somehow Kalafat survived floods for over a century or more.  If it was not saved from flood destruction year after year by being on a hill, then perhaps the mounds around it protected it.  In other words, perhaps the ancient flood control works in the plain protected the old location of the village.